This invention relates generally to systems for automatically testing fluid specimens, e.g. urine or other body fluids, to detect chemical substances or components therein. Such systems can be used, for example, to screen employee applicants for illegal drug use.
Employee drug testing typically involves an initial screening test to identify specimens which are negative (i.e., no drugs present). This test is usually performed with a low cost immunoassay which is very sensitive to small quantities of drug metabolites. If a drug metabolite is detected (referred to as xe2x80x9cpresumptive positivexe2x80x9d), the specimen is then subjected to a confirmation test which typically utilizes a highly specific test method, such as a gas chromatography/mass spectrometry (GC/MS), to identify the specific drug components in the specimen.
Traditionally, both the screening and confirmation tests were performed in a common facility, i.e., centralized laboratory. More recently, systems have been implemented which perform the screening test at a local service site. They typically employ drug test kits which follow the tradition of home pregnancy test kits, i.e., they detect the presence of a specific drug substance(s) in a urine specimen. Such drug test kits generally identify, in human readable form, the drug(s) being tested to indicate the presence (or absence) of that drug. The screening test result with respect to each particular drug, and to the specimen as a whole, can either be (1) negative or (2) presumptively positive. If presumptively positive, then the specimen is generally sent to a remote laboratory for confirmation testing.
Inasmuch as disclosure of a presumptive positive test result can adversely impact an applicant""s record and could potentially lead to litigation and employer liability, it is extremely important to prevent the inadvertent dissemination of test result data.
The aforementioned parent U.S. Pat. No. 6,342,183 describes an apparatus for locally analyzing a specimen to qualitatively detect specified chemical components therein. The apparatus includes an assaying device comprised of a cup for collecting a fluid specimen and a cap carrying at least one test strip for visually reacting to one or more specified chemical components in the specimen. The assaying device is preferably configured to interact with a reader device capable of reading the reaction of the test strip to produce an electronic data output.
More particularly, the apparatus described in Application Ser. No. 09/245,175 includes an open cup defining an interior volume for accommodating a fluid specimen and an attachable cap configured for mounting on the cup to seal the interior volume. The cap carries at least one test strip and an integrated aliquot delivery mechanism actuatable to wet the test strip with an aliquot derived from the fluid specimen. The aliquot delivery mechanism preferably comprises a pump in the form of a plunger for forcing an aliquot of the fluid specimen onto the test strip. The plunger can be actuated either manually or automatically, e.g., by a piston controlled by a compatible reader device. The reader device preferably includes a microprocessor based controller for actuating the aliquot delivery mechanism, a camera for producing an image of the test strip, and a processor for analyzing the image to produce test result data. The test result data, along with identification data read from a label carried by the cap, can then be stored or communicated, e.g., via a modem.
The present invention is directed to an improved system and components thereof for automatically testing a fluid specimen, e.g. urine, saliva, or other body fluids, to indicate for the presence of specified chemical components in the specimen.
A system in accordance with the invention preferably utilizes an assaying device comprised of a collection cup and a cap which carries at least one test strip. The device includes an integrated aliquot delivery mechanism actuatable to wet the test strip with an aliquot derived from the fluid specimen. The assaying device is configured to operate in conjunction with an electronic reader device capable of actuating the aliquot delivery mechanism and reading the reaction of the test strip.
A preferred reader device in accordance with the invention preferably defines a keyed receptacle for accommodating a complementary shaped cup housing in a particular orientation. The reader device is comprised of a camera for capturing the image of a test strip, an actuator for actuating an aliquot delivery mechanism, and a microprocessor/controller for (1) controlling the camera and actuator and (2) processing the image.
In a preferred embodiment of the invention, the reader device preferably also includes a network connectivity device, e.g. modem, for enabling communication with a remote host computer. Although each reader device can operate independently as a stand-alone device, a preferred system in accordance with the inventor employs a host computer or server, which communicates, via a public and/or private network, with a plurality of reader devices located at separate service sites. Each service site can be configured to operate as a xe2x80x9cthin clientxe2x80x9d with primary control being exercised by the host computer via the network. Alternatively, primary control can be exercised by the reader device at each site with only high level supervisory control coming from the host computer.
A preferred assaying device in accordance with the invention includes a cap carrying multiple test strips including at least one component test strip and at least one adulteration test strip. The cap is either formed of transparent material or is provided with transparent windows to permit external viewing of the test strips by the reader device camera. The cap preferably also carries one or more fiducial marks to facilitate image processing. Further, the cap preferably also carries machine readable identification information, e.g., a bar code label, to positively associate the specimen and test results with the correct individual. Preferably, the cap does not bear any human readable indicia identifying the specimen donor or indicating test results.
A preferred test strip for testing for the presence of specific chemical components is configured with multiple latent lines (i.e., markings) which can become visible when the strip is wetted. The lines preferably include a control or reference line and multiple drug lines each related to a different chemical component. If all of the latent lines visually appear within a certain test interval, e.g., up to eight minutes, after the strip has been wetted, this will indicate the absence of the specific chemical components sought. However, if any of those specific chemical components are present in concentrations above a certain threshold, their presence will suppress the appearance of one or more of the drug lines to indicate the presence of such chemical components.
A preferred reader device in accordance with the invention includes a camera located so that the cap is imaged onto the camera focal plane. The reader device includes a piston motor for driving a piston against the assaying device to deliver an aliquot to the test strips. The piston motor also moves a light shield into place around the cap enabling a light source to illuminate the cap to enhance the image for the camera.
A preferred automatic testing system in accordance with the invention operates as follows:
1. Fluid deposited into cup at local site; secure cap in tamper evident fashion;
2. Site administrator places assaying device, i.e., cup and cap, into xe2x80x9ckeyedxe2x80x9d receptacle of local reader and enters ID information;
3. Reader alerts host computer via communication network;
4. Initiate automatic reader operational sequence:
a. capture cap image and verify acceptability to proceed
b. run piston motor to advance piston into assaying device to force fluid up channels to wet component and adulteration test strips
c. capture cap image and verify acceptability to proceed
d. periodically capture additional cap images during development interval up to about eight minutes
e. analyze captured image data to determine
1. test validity
2. test results
f. locally display test validity/results and/or communicate test validity/results to Host computer
g. run piston motor to withdraw piston from cup
5. Site administrator removes cup from reader
The camera produces a digital representation of the image incident on the camera focal plane. The processor then analyzes the digital representation to determine the color of the adulteration test strip and to locate visible markings on the component test strip coincident with the reference and drug lines. Image analysis is preferably performed by initially using fiducial marks on the cap to first precisely locate the cap image relative to a reference image. This can, for example, involve rotating, translating, and/or scaling the cap image. Thereafter, the digital representation of each test strip is examined to determine the presence (or absence) of drug lines. This involves first locating the strip reference line by effectively xe2x80x9cdrawingxe2x80x9d a rectangular region around the reference line. The region can be considered as a rectangular matrix of pixels having rows extending across the strip width, each row being comprised of multiple column positions. For each row, the image is examined to determine whether the pixel at each column position exceeds a threshold. The sum of pixels exceeding the threshold is determined for each row. These row sums produce a graph whose x axis is related to the height (i.e., number of rows) of the region and whose y axis is related to the values of the individual row sums. A bell shaped curve will result whose peak locates the reference line. If no reference line is located, the test is terminated. If the reference line is located, then the examination continues in order to locate the drug lines. The detection of drug lines is more difficult because the amplitude, i.e., dark or light, of its pixels can vary widely dependent on several factors including wetting uniformity, urine color, variations amongst test strips, exposure time, etc. In order to compensate for these variations, each drug region is preferably divided into left, center, and right portions. A drug line is presumed to occupy the center portion of each region. However, its exact position and exact width can vary attributable to the aforementioned factors. Moreover, its brightness difference in relation to neighboring areas can be very subtle. Hence, a procedure is used to determine the weight of a line on a relative basis.
For example, for each drug region graph, the total area under each of three regions (left, center, right) is calculated. The left and right region areas are then numerically summed, and this resulting total area sum is multiplied by an experimentally determined xe2x80x9cweighting value,xe2x80x9d thus producing a weighted sum. If the area of the center region is less than or equal to the weighted sum, no line is present. By using urine samples with known drug concentrations, a weighting value of 0.75 has been experimentally determined to produce very acceptable results.